Cannabis for MS Spasticity: A Systematic Review With Critically Unreliable Pooled Estimates

Nine trials encompassing 2,544 patients were pooled to assess cannabis-based therapies for multiple sclerosis spasticity, but extreme statistical heterogeneity reaching 100%, numerically implausible effect sizes, and suspected publication bias fundamentally undermine the review’s headline conclusions and render its pooled estimates uninterpretable for clinical decision-making.

Why This Matters

Spasticity affects up to 80% of people living with multiple sclerosis and remains one of the most treatment-resistant symptoms driving disability and reduced quality of life. Cannabis-based medicines, including nabiximols and whole-plant extracts, are increasingly prescribed or self-administered for this indication, and several countries have created regulatory pathways specifically for MS spasticity. A methodologically sound synthesis of clinical trial data would directly shape prescribing guidelines and reimbursement decisions, making it essential that clinicians understand not only what this meta-analysis claims but also why its quantitative conclusions cannot be taken at face value.

Clinical Summary

Cannabis-based therapies have attracted sustained clinical interest for MS spasticity because both THC and CBD modulate endocannabinoid signaling pathways implicated in motor tone regulation. This systematic review and meta-analysis, published in Clinical Therapeutics in 2026 and conducted according to PRISMA guidelines, searched PubMed, Cochrane Library, and Google Scholar through October 2024, ultimately pooling data from nine clinical trials conducted between 2003 and 2021. The included studies evaluated heterogeneous formulations, ranging from whole-plant extracts and oils to smoked cannabis containing THC and CBD in varying ratios, and measured spasticity using at least three distinct instruments: the Ashworth Scale, the Numerical Rating Scale, and the Visual Analog Scale.

The review reported an overall pooled mean difference of 39.19 (95% CI: 34.32 to 44.05) favoring cannabis-based therapies, with subgroup analyses showing a short-term mean difference of 4.53 (crossing zero) and a long-term mean difference of 75.81. However, the overall I-squared statistic reached 100%, indicating that virtually all observed variance was between-study rather than within-study, and the overall mean difference of 39.19 is numerically impossible on any standard spasticity scale (the Ashworth Scale ranges 0 to 4; NRS typically 0 to 10), strongly suggesting inappropriate cross-scale pooling. Funnel plot asymmetry raised additional concerns about publication bias. The authors acknowledged the heterogeneity but still concluded that cannabis-based therapies produce “clinically meaningful improvements,” a characterization the data as presented cannot support. Adverse events, including dizziness and dry mouth, were described as generally mild, though the safety evidence base across only nine heterogeneous trials is insufficient for confident conclusions. The authors noted that further high-quality, standardized trials are needed before firm clinical recommendations can be made.

Dr. Caplan’s Take

I see patients with MS spasticity regularly, and many of them arrive having already read headlines suggesting that cannabis is proven to help. This review asks exactly the right question, and I appreciate the effort to synthesize the trial literature. But the numbers simply do not hold together. A pooled mean difference of 39 on scales that max out at 4 or 10 tells us something went wrong in the analysis, not that cannabis is producing enormous effects. When a patient asks me about this study, I have to be honest: the individual trials do generally lean positive, but this particular synthesis cannot quantify how positive.

In practice, I consider cannabis-based therapies, particularly standardized products like nabiximols, as part of a broader spasticity management plan for patients who have not responded adequately to first-line agents such as baclofen or tizanidine. I discuss realistic expectations, monitor for cognitive and mood effects that matter especially in MS, and use validated scales consistently to track response. I do not cite this meta-analysis as evidence of efficacy because its pooled estimates are not interpretable, but I also do not dismiss the signal from individual trials that a subset of patients may benefit.

Clinical Perspective

This review sits at a frustrating point in the research arc for cannabis and MS spasticity. Several well-conducted individual trials, most notably the CAMS and MUSEC studies, have demonstrated modest but real benefits of cannabinoid preparations on patient-reported spasticity. What this meta-analysis fails to accomplish is the next logical step: generating a reliable pooled effect estimate that could anchor guideline recommendations. The I-squared of 100% is not a minor statistical footnote; it means the included studies are so dissimilar in populations, interventions, comparators, and outcome instruments that combining them yields a number with no clinical meaning. Clinicians should not quote the reported mean differences in patient conversations or documentation.

From a pharmacological standpoint, THC-containing preparations carry particular considerations in MS patients, including potential exacerbation of fatigue, cognitive impairment, and mood disturbance, symptoms that already burden this population. Drug interactions with disease-modifying therapies, particularly those metabolized via CYP3A4 and CYP2C19 pathways, warrant monitoring. Clinicians managing spasticity in MS should continue to rely on established first-line agents and consider standardized cannabinoid products as adjunctive therapy only in refractory cases, using consistent validated outcome measures to assess individual response over defined trial periods of four to six weeks.

Study at a Glance

Study Type

Systematic review and meta-analysis

Population

Adults with multiple sclerosis and spasticity (2,544 patients across 9 trials)

Intervention

Cannabis-based therapies including whole-plant extracts, oils, and smoked cannabis (THC and/or CBD)

Comparator

Placebo or standard care (variable across trials)

Primary Outcomes

Spasticity measured by Ashworth Scale, Numerical Rating Scale, and Visual Analog Scale

Sample Size

9 trials, 2,544 patients

Journal

Clinical Therapeutics

Year

2026

DOI or PMID

Not available in source material

Funding Source

Not reported in available text

What Kind of Evidence Is This

This is a peer-reviewed systematic review and meta-analysis, a study type that ordinarily sits near the top of the evidence hierarchy because it synthesizes multiple primary studies into a single quantitative estimate. However, a meta-analysis is only as reliable as its component studies and the appropriateness of combining them. The I-squared value of 100% for the primary outcome means the pooled estimate reflects virtually no common signal across the included trials, reducing this synthesis to a collection of individual study results that cannot be meaningfully combined into a single effect size.

How This Fits With the Broader Literature

The broader literature on cannabis-based therapies for MS spasticity includes several landmark trials. The CAMS study (Zajicek et al., 2003) found no significant effect on the Ashworth Scale but reported patient-perceived improvement, while the MUSEC trial (Zajicek et al., 2012) demonstrated a modest benefit on a category rating scale. Regulatory approval of nabiximols (Sativex) in multiple countries was based on a series of enriched-enrollment randomized withdrawal trials showing NRS reductions of approximately 1 to 2 points. The current review’s qualitative finding that individual trials lean modestly positive is consistent with this established literature. What it adds, unfortunately, is not a more precise pooled estimate but rather a demonstration of how difficult it remains to synthesize trials that use different formulations, doses, durations, and outcome instruments into a single number.

Common Misreadings

The most likely overinterpretation is treating the reported mean difference of 39.19 as evidence that cannabis-based therapies produce large, clinically meaningful improvements in MS spasticity. This number almost certainly results from inappropriate aggregation of raw scores across scales with fundamentally different ranges and units, producing a composite statistic that corresponds to no real-world clinical measurement. A second common misreading would be interpreting the long-term subgroup finding (MD 75.81) as evidence that cannabis works better with prolonged use. Given the maximal heterogeneity and the mathematical impossibility of these values on any standard spasticity scale, this finding more likely reflects analytic artifact than a genuine duration-response relationship.

Bottom Line

This meta-analysis addresses a clinically important question but produces pooled estimates that are numerically implausible and statistically uninterpretable due to extreme heterogeneity and likely inappropriate cross-scale pooling. It does not establish the magnitude of benefit from cannabis-based therapies for MS spasticity. Clinicians should continue to rely on individual trial data and existing guideline recommendations, reserving standardized cannabinoid preparations as adjunctive therapy in refractory cases with careful individual outcome monitoring.

References

1. Zajicek J, Fox P, Sanders H, et al. Cannabinoids for treatment of spasticity and other symptoms related to multiple sclerosis (CAMS study): multicentre randomised placebo-controlled trial. Lancet. 2003;362(9395):1517-1526. doi:10.1016/S0140-6736(03)14738-1
2. Zajicek JP, Hobart JC, Slade A, Barnes D, Mattison PG; MUSEC Research Group. Multiple sclerosis and extract of cannabis: results of the MUSEC trial. J Neurol Neurosurg Psychiatry. 2012;83(11):1125-1132. doi:10.1136/jnnp-2012-302468
3. Novotna A, Mares J, Ratcliffe S, et al. A randomized, double-blind, placebo-controlled, parallel-group, enriched-design study of nabiximols (Sativex), as add-on therapy, in subjects with refractory spasticity caused by multiple sclerosis. Eur J Neurol. 2011;18(9):1122-1131. doi:10.1111/j.1468-1331.2010.03328.x
4. Collin C, Ehler E, Waberzinek G, et al. A double-blind, randomized, placebo-controlled, parallel-group study of Sativex, in subjects with symptoms of spasticity due to multiple sclerosis. Neurol Res. 2010;32(5):451-459. doi:10.1179/016164109X12590518685660
5. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. doi:10.1136/bmj.n71